Synthesis and Testing of Mixtures

Some combinatorial synthetic approaches rely on the synthesis and testing of mixtures to achieve high throughput. Typically, one wants to maximize the diversity of each pool, so the activity of a pool is indicative of the activity of its most potent compound. However, when sensitivity of the test or solubility of the compounds has been a problem, some groups have, instead, pooled similar compounds, amplifying a weak signal by the epitope effect, that is, the cumulative activity of many related compounds. Regardless of whether compounds are pooled, the methods described previously allow one to maximize either the similarity or diversity of the pools. 

Burgess et al. used both computer simulations and an algebraic expression based on the Poisson distribution to analyze the number of beads required to have confidence that every intended compound is present in the library. For typical library sizes, if the number of beads is an order of magnitude greater than the total number of compounds in the library, every compound should be present on at least one bead. Zhao et al. used the Pearson statistic to determine the number of beads needed to be confident that either the smallest individual error or overall relative error in concentration is less than a given threshold. 

There are several ways to deconvolute or identify the individual active compounds in an active mixture. These include iterative resynthesis and the synthesis of overlapping mixtures, such as the positional scanning method of Houghten et al. In iterative resynthesis, the most active pools are resynthesized as several smaller subpools with a variable position resolved. This process is repeated until every position has been resolved and individual active compounds have been made. In Houghten s method, the complete library is 

Screening combinatorial libraries should not imply a random or irrational approach to drug discovery. Far from obviating the use of computers, the increased complexity of combinatorial drug discovery makes their use ever more essential. Although the number of compounds that could be made is enormous, a well-planned experimental design selects a reasonable number of compounds that sample the range of accessible properties. 

315 (1993). Recent Advances in the Generation of Molecular Diversity. See also R. Baum, Chem. Eng. News, February 12,1996, p. 28. Combinatorial Chemistry. S. Borman, Chem. Eng. News, February 12,1996, p. 29. Combinatorial Chemists Focus on Small Molecules, Molecular Recog- 

---